The fact that traumatic brain injury (TBI) is a major cause of long-term disability for young adults underscores the need to conceptualize coping with TBI as a life-long process of recovery and rehabilitation. Successful characterization of neuropathologic markers over the evolving course of recovery will become an essential step toward developing more effective prognostication and intervention. While it is well established that behavioral outcome measures show significant improvements over months and even years after the injury, the underlying neural mechanisms of such recovery are not well understood. With the recent advances of in vivo neuroimaging, it may now be possible to more accurately measure the initial injury pathology and to shed light on the mechanisms of the natural recovery process in a longitudinal design. However, challenges in image analyses have hampered TBI researchers from taking full advantage of new imaging modalities. Several major data analysis issues in TBI neuroimaging research are: 1) quantifying the consequences of diffuse and focal injuries, 2) registering heterogeneous brains into a common space, and 3) assessing structure-function relationship (e.g., diaschisis). We have recently demonstrated the feasibility of applying high-dimensional large deformation image registration method in solving these methodological challenges, identifying the regions showing structural and functional group differences between controls and TBI with unprecedented precision. Based on these preliminary work, the current project aims 1) to determine the pattern of longitudinal changes in structural and functional neuroimaging indices associated with moderate to severe diffuse axonal injury and their relationship with behavioral improvements, 2) to develop a neuropathologically based injury severity measure using longitudinal changes in the early post-acute phase, and 3) to develop a structure-function discrepancy index, based on the difference between the integrity of structural and functional imaging measures obtained at different points of post-injury, and collect preliminary data on its relationship with behavioral recovery. We propose to evaluate 60 individuals with moderate to severe diffuse TBI in the post-acute phase with structural (gross volume change and white matter integrity) and functional (cerebral perfusion and functional connectivity) neuroimaging measures at multiple time points (3, 6, and 12 months from the date of injury). Thirty six demographically matched healthy control subjects will be evaluated for comparison purposes. To measure longitudinal improvements in behavior, a global behavioral outcome measure and a neuropsychological test battery consisting of five executive function tests will be administered at each time point. PUBLIC HEALTH RELEVANCE: PROJECT NARRATIVE Taking advantage of our innovative neuroimaging methods, this project will track the changes in structure and function of the brains in individuals with traumatic brain injury. These changes will be related to concurrent behavioral improvement. Results from this project will greatly benefit future research aiming to investigate the neural mechanisms of behavioral recovery, ultimately contributing the development of efficient intervention.